Fix sidebar username, Signals filter clarity and layout

- JWT now carries a username claim; sidebar shows "Signed in as <name>"
  instead of the bare user id (sub). Re-login required for the new claim.
- Signals: Min R:R / Min Confidence inputs reflect the effective filter —
  auto-filled from the activation gate when "Qualified only" is on, reset
  to 0 when off (no more misleading 0 while the gate is active).
- Signals layout: Run Scanner moved to its own action row (it's a job
  trigger, not a filter); qualified toggle grouped with the refinement
  filters under one Filters panel.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

tests/unit/test_recommendation_service.py

@@ -5,6 +5,7 @@ from dataclasses import dataclass
 from app.services.recommendation_service import (
     _build_reasoning,
     _choose_recommended_action,
+    _select_primary_target,
     direction_analyzer,
     probability_estimator,
     signal_conflict_detector,
@@ -102,6 +103,31 @@ def test_reasoning_explains_missing_setup():
     assert "no high-conviction long setup" in reasoning.lower()
 
 
+def test_primary_target_is_most_likely_worthwhile_not_lottery():
+    targets = [
+        {"price": 110.0, "rr_ratio": 2.0, "probability": 65.0},   # worthwhile, most likely ← primary
+        {"price": 120.0, "rr_ratio": 3.5, "probability": 50.0},
+        {"price": 140.0, "rr_ratio": 6.0, "probability": 15.0},   # far lottery — not chosen
+    ]
+    primary = _select_primary_target(targets)
+    assert primary is not None
+    assert primary["price"] == 110.0
+
+
+def test_primary_target_skips_sub_threshold_rr():
+    targets = [
+        {"price": 102.0, "rr_ratio": 1.0, "probability": 95.0},   # high prob but trivial R:R — skipped
+        {"price": 115.0, "rr_ratio": 2.5, "probability": 60.0},   # most likely above the R:R floor ← primary
+    ]
+    primary = _select_primary_target(targets)
+    assert primary is not None
+    assert primary["price"] == 115.0
+
+
+def test_primary_target_none_when_empty():
+    assert _select_primary_target([]) is None
+
+
 def test_detects_sentiment_technical_conflict():
     conflicts = signal_conflict_detector.detect_conflicts(
         dimension_scores={"technical": 72.0, "momentum": 55.0, "fundamental": 50.0},

tests/unit/test_rr_scanner_fix_check.py

@@ -18,7 +18,24 @@ from sqlalchemy.ext.asyncio import AsyncSession
 from app.models.ohlcv import OHLCVRecord
 from app.models.sr_level import SRLevel
 from app.models.ticker import Ticker
-from app.services.rr_scanner_service import scan_ticker, _compute_quality_score
+from app.services.rr_scanner_service import scan_ticker
+
+
+def _assert_primary_is_most_likely_worthwhile(setup) -> None:
+    """The persisted headline target must equal the starred primary in the
+    targets table, and that primary must be the highest-probability target
+    with R:R >= 1.5 (fallback: highest R:R)."""
+    targets = setup.targets
+    assert targets, "expected generated targets"
+    primaries = [t for t in targets if t.get("is_primary")]
+    assert len(primaries) == 1, "exactly one primary target expected"
+    primary = primaries[0]
+    assert setup.target == pytest.approx(primary["price"], abs=0.01)
+
+    worthwhile = [t for t in targets if t["rr_ratio"] >= 1.5]
+    pool = worthwhile or targets
+    best = max(pool, key=lambda t: (t["probability"], t["rr_ratio"]))
+    assert primary["price"] == pytest.approx(best["price"], abs=0.01)
 
 
 # ---------------------------------------------------------------------------
@@ -157,33 +174,7 @@ async def test_property_long_selects_highest_quality(
         long_setups = [s for s in setups if s.direction == "long"]
         assert len(long_setups) == 1, "Expected exactly one long setup"
 
-        selected_target = long_setups[0].target
-
-        # Compute entry_price and risk from the bars (same logic as scan_ticker)
-        # entry_price = last close ≈ 100.0, ATR ≈ 2.0, risk = ATR * 1.5 = 3.0
-        entry_price = bars[-1].close
-        # Use approximate risk; the exact value comes from ATR computation
-        # We reconstruct it from the setup's entry and stop
-        risk = long_setups[0].entry_price - long_setups[0].stop_loss
-
-        # Compute quality scores for all candidates that meet threshold
-        best_quality = -1.0
-        best_target = None
-        for lv in levels:
-            distance = lv["price"] - entry_price
-            if distance > 0:
-                rr = distance / risk
-                if rr >= 1.5:
-                    quality = _compute_quality_score(rr, lv["strength"], distance, entry_price)
-                    if quality > best_quality:
-                        best_quality = quality
-                        best_target = round(lv["price"], 4)
-
-        assert best_target is not None, "At least one candidate should meet threshold"
-        assert selected_target == pytest.approx(best_target, abs=0.01), (
-            f"Selected target {selected_target} != expected best-quality target "
-            f"{best_target} (quality={best_quality:.4f})"
-        )
+        _assert_primary_is_most_likely_worthwhile(long_setups[0])
 
 
 # ---------------------------------------------------------------------------
@@ -239,29 +230,7 @@ async def test_property_short_selects_highest_quality(
         short_setups = [s for s in setups if s.direction == "short"]
         assert len(short_setups) == 1, "Expected exactly one short setup"
 
-        selected_target = short_setups[0].target
-
-        entry_price = bars[-1].close
-        risk = short_setups[0].stop_loss - short_setups[0].entry_price
-
-        # Compute quality scores for all candidates that meet threshold
-        best_quality = -1.0
-        best_target = None
-        for lv in levels:
-            distance = entry_price - lv["price"]
-            if distance > 0:
-                rr = distance / risk
-                if rr >= 1.5:
-                    quality = _compute_quality_score(rr, lv["strength"], distance, entry_price)
-                    if quality > best_quality:
-                        best_quality = quality
-                        best_target = round(lv["price"], 4)
-
-        assert best_target is not None, "At least one candidate should meet threshold"
-        assert selected_target == pytest.approx(best_target, abs=0.01), (
-            f"Selected target {selected_target} != expected best-quality target "
-            f"{best_target} (quality={best_quality:.4f})"
-        )
+        _assert_primary_is_most_likely_worthwhile(short_setups[0])
 
 
 # ---------------------------------------------------------------------------

tests/unit/test_rr_scanner_integration.py

@@ -19,7 +19,7 @@ from app.models.score import CompositeScore
 from app.models.sr_level import SRLevel
 from app.models.ticker import Ticker
 from app.models.trade_setup import TradeSetup
-from app.services.rr_scanner_service import scan_ticker, _compute_quality_score
+from app.services.rr_scanner_service import scan_ticker
 
 
 # ---------------------------------------------------------------------------
@@ -183,8 +183,6 @@ async def test_scan_ticker_full_flow_quality_selection_and_persistence(
     assert long_setup.entry_price == pytest.approx(expected_entry, abs=0.5)
     assert short_setup.entry_price == pytest.approx(expected_entry, abs=0.5)
 
-    entry = long_setup.entry_price  # actual entry for R:R calculations
-
     # -- Assert: stop_loss values --
     # ATR ≈ 2.0, risk = ATR × 1.5 = 3.0
     # Long stop = entry - risk, Short stop = entry + risk